Around September 2018

​
​

What we were doing as the storage team at Scaleway was a product of both the context and the time we were in.

We were almost acting as a startup within an incubator. We did whatever was necessary to move on. Launching in Amsterdam instead of Paris first was one of the examples. We wanted the product to be live so customers could use it. There was a delay in the work on DC5 so we just “pivoted” to Amsterdam.

Every scope increase we did was to be able to move faster or reduce risk.

All the steps that I narrate in this story about Object Storage were also happening on block storage, albeit with a slight lag. This was led internally by Florian. Théo and I supervised.

Around that time, September 2018, Scaleway’s marketing department was preparing a naming strategy. The hype of the moment was space, so they prepared a strategy to name products around space themes. After some thought we said that object storage should be named Oort cloud and block storage asteroid belt. Why? Because that reflected the latency and the scale. Sadly, for me at least, the marketing team pivoted away from space metaphors and into elements.

The push forward, not waiting for other teams, was what made us move so fast to be ready. In October 2018, 11 months after the pivot, we launched the private beta of object storage in Amsterdam. We were the first team to launch a service in beta after the pivot.

A lot of that speed came from our combined experience. By carefully splitting and sequencing the work we had moved in parallel and gained a lot of precious time.

One example of this was to create an administrative service API to handle trust & safety requests. Implemented by Nicolas, this made it possible to handle administrative requests.

Even though we had launched public beta so fast, I don’t remember celebrating the launch. I do remember looking at the many more things we had to do. How soon we needed to build them. Not celebrating the release to beta was a mistake. We had accomplished so much. The work so far had a huge value. But because it was not complete, I didn’t think it was time to celebrate yet. But, the work never ends, does it? Now I know you need to celebrate milestones, otherwise it eventually only feels like a burden.

A question you might be wondering is: weren't there people leading from the product side? Decision made from above? And that is a good question. There wasn’t at the time. We did what we had to do. Later, more layers and steps were added but at this precise time there wasn’t a lot of guidance. This fared well for some teams, like us and the database team led by Nicolas Limage. It fared less for other teams who needed more guidance.

For us, storage, it was a blast. The freedom and the weight. We were ambitious. We pushed everything forward.

Part of the credit of our success is due to OpenIO. In October 2018, an event was organized at la Maison. It was the OpenIO festival.

​
​

(1) Private preview in September 2018. https://x.com/Scaleway/status/1039529326461562881

(2) https://www.space.com/40601-space-station-astronauts-spacewalk-expedition-55-webcast.html

If you have missed it, you can read the previous episode here​

To pair with :

• Wallowa lake monster - Sufjan Stevens
• Aama by Frederik Peeters